D

LJa

b

ARR2AA

KEDDKM

I

alatdtelHa(il

siM

2

0h

Immunobiology 218 (2013) 1019– 1025

Contents lists available at SciVerse ScienceDirect

Immunobiology

jou rn al hom ep age: www.elsev ier .com/ locate / imbio

O�−�+ expression in favor of HLA-DR engagement in exosomes

ina Papadimitrioua,1, Ioanna Zervaa,1, Mirella Georgouli a,1, Takis Makatounakisb,oseph Papamatheakisa,b, Irene Athanassakisa,∗

Department of Biology, Laboratory of Immunology, University of Crete, Heraklion 714 09, Crete, GreeceInstitute of Molecular Biology and Biotechnology, Foundation for Research and Technology – Hellas, Heraklion 70013, Crete, Greece

a r t i c l e i n f o

rticle history:eceived 29 July 2012eceived in revised form4 December 2012ccepted 29 December 2012vailable online 7 January 2013

eywords:xosomes

a b s t r a c t

The expression of DO� and not DO�, in addition to the high intracellular DR, low DM levels and absence ofsurface DR expression in K562 and HL-60 cells introduce alternative regulatory pathways in DR traffickingand consequently the antigen presentation process. The present study attempted to define the naturallyoccurring DO� negative state and explain the role of DO� in the intracellular DR accumulation in K562 andHL-60 cells. Despite the absence of DO�, the DO� chain was detected in the endosomal compartments.The lack of DO� was found to be partially responsible for the absence of DR from the cell membranesince stable K562-DO� transfectants allowed expression of membrane DR. This expression could besignificantly increased upon DM induction by IFN-�, indicating that DM was another limiting factor

O� lacking expressionO� expressing cells562 and HL-60 leukemic cellsembrane HLA-DR expression

for the migration of DR to the cell surface of K562 and HL-60 cells. Furthermore, intracellular DR co-localized with the exosome specific marker CD9, while culture supernatants were shown to containexosome-engaged and exosome free DR activity as evaluated by SDS-page followed by western blot,ELISA and transmission electron microscopy analysis. These findings indicated that in DO�−�+ cells, DRmolecules were programmed to secretion rather than surface expression. The presented results providenovel regulatory processes as to DR trafficking, avoiding expression to the cell surface.

ntroduction

Non-classical HLA-DO (DO) is believed to negatively regulate thentigen presentation process in B cells, trophoblasts, thymic epithe-ial cells and possibly other not yet defined cell types, providing andditional mechanism for the maintenance of a tolerogenic stateo the organism (Denzin et al. 1997). DO expression has also beenetected in many cases of primary acute myeloid leukemia, wherehe DO:DM ratio correlated with the CLIP:DR ratio (Chamuleaut al. 2004). Previous studies in the leukemic K562 and HL-60 cellines have shown that although these cells do not express surfaceLA-DR (DR), they contain intracellular pools of these moleculesnd express high levels of DO� and CD74, but not DO� chains

Papadimitriou et al. 2008). It has thus been postulated that DOs part of the escape mechanism from immune surveillance ineukemic cells.

Abbreviations: K-DO�, stable transfectants of K562 cells with DO�; K-DO�,table transfectants of K562 cells with DO�; mbDR, membrane HLA-DR; Intradr,ntracellular HLA-DR; DR, HLA-DR; DM, HLA-DM; DO, HLA-DO; IFN-�, interferon-�;

HCII, major histocompatibility class II.∗ Corresponding author at: Department of Biology, University of Crete, P.O. Box

208, 714-09 Heraklion, Crete, Greece. Tel.: +30 2810394355; fax: +30 2810394379.E-mail address: athan@biology.uoc.gr (I. Athanassakis).

1 Equal contribution.

171-2985/$ – see front matter © 2013 Elsevier GmbH. All rights reserved.ttp://dx.doi.org/10.1016/j.imbio.2012.12.003

© 2013 Elsevier GmbH. All rights reserved.

DO is known to be a lysosomal-resistant protein with low poly-morphism, expression limited to specific cell types, regulatingimmune response by controlling the availability of DM (Denzinet al. 2005; vam Ham et al. 1997; vam Ham et al., 2000; Chenet al. 2002). On the other hand, DM is catalyzing the release of CLIPand loading of antigenic peptide to DR, is present in all antigenpresenting cells (APCs), is lysosome resistant, shows limited poly-morphism and does not bind antigenic peptides (Morris et al. 1994;Kropshofer et al. 1996). The activity of DO is pH-dependent and hasbeen postulated to exert its inhibitory activity to DM in the earlycompartments of the endocytic pathway by limiting the pH rangein which DM is active (vam Ham et al. 2000). Recently, DO has beenpostulated to impair the incorporation of DM into exosomes, pro-moting thus DR excretion through exosomes (Xiu et al. 2011). Therestricted cellular distribution of DO, implies different regulatorycontrol mechanisms as compared to classical MHC class II (MHCII)antigens.

Although DO� has been shown to co-regulate with other MHCIImolecules (similar enhancing elements in the promoter sequencesand inducibility by interferon-� (IFN-�), DO� is subjected to dif-ferent regulatory mechanisms and is not stimulated by IFN-�

(Tonnelle et al. 1985; Wake and Flavell 1985), while its activ-ity seems to depend on a di-leucine cytoplasmic motif, which isa sorting signal regulating trafficking and endocytosis (Xiu et al.2011). However, it has been shown that mutation of the di-leucine

1 unobio

sw

roloegortcTnDlbtf

M

R

R(BptCmicdua(Hba

I

(caHasIeewr

N

mpcuw

020 L. Papadimitriou et al. / Imm

equence in DO� did not alter its lysosomal sorting when associatedith DM molecules (Brunet et al. 2000).

The differential role of DO� and DO� seems to play an importantole in DR trafficking. H2-Oa – deficient mice generated by homol-gous recombination have been shown to express physiologicalevels of classical MHCII molecules on APCs, while in the absencef H2-M these were not loaded with antigenic peptide (Liljedahlt al. 1998). Other studies, using various HeLa transfectants sug-ested that association of DO with DM is required for efficient exitf DO from the ER (Samaan et al. 1999), while additional studieseport that although DO� contains functional sorting motifs, theight compartmentalization of HLA-DO observed inside B lympho-ytes is controlled by the DO� chain and DM (Brunet et al. 2000).he present study was designed to exploit the role of DO� in theegative expression of surface DR and correlate its expression toR excretion. Using the K562 and HL-60 cell lines which naturally

ack DO�, express high levels of DO� and low levels of DM, it coulde possible to exploit the mechanisms disallowing DR migrationo the cell membrane, which could be thereafter linked to escaperom immune surveillance.

aterials and methods

eagents

Uncoupled mouse anti-HLA-DP, DQ, DR (HB145; ATCC,ockville, Maryland, USA) or FITC-labeled mouse anti-HLA-DRL243,Santa Cruz, CA, USA), mouse anti- HLA-DO (DoB.L1, BDiosciences, San Diego, California, USA), anti-rab7 (C19, affinityurified goat polyvalent antibody, Santa Cruz) as well as coupledo PE mouse anti-HLA-DM (MaPDM1, BD Biosciences) and anti-D9-PE conjugated (ImmunoTools GmbH, Friesoythe, Germany)onoclonal antibodies were used at the concentration of 1 �g/ml

n immunofluorescence experiments. Anti-mouse IgG Fab fragmentoupled to FITC was used as secondary antibody for HLA-DO�etection and anti-goat IgG Fab fragment coupled to PE wassed as secondary antibody for rab7 detection. Anti-mouse andnti-goat IgG Fab fragment coupled to horseradish peroxidaseSigma–Aldrich Co., MO, USA) were used as secondary antibody forLA-DO� and rab7 detection in ELISA experiments. Human recom-inant IFN-� was purchased from Endogen (Cambridge, MA, USA)nd used at the concentrations of 1500 pg/ml.

mmunofluorescence experiments

Immunofluorescence was performed as previously describedPapadimitriou et al. 2008). For the intracellular staining, theells were fixed with ice-cold 4% paraformaldehyde for 5 minnd permeabilized using a HBSS–saponin solution (HBSS, 0.01 MEPES, 0.3% saponin). Fluorescence intensity was evaluated using

FACS analysis system (Calibur Flow Cytometer, Becton Dickin-on). Where appropriate, the negative controls included mousegG (Sigma, 1 �g/ml) or control buffer (PBS-BSA). Expression lev-ls were evaluated using the histogram subtraction tool of FCSxpress® software. For confocal microscopy analysis the cellsere fixed using 25 �l/ml Mowiol (Sigma). All experiments were

epeated at least 3 times.

ative and transfected cell lines

HL-60, K562 and Raji cells were purchased from ATCC andaintained in RPMI culture medium (Gibco, Grand Island, NY) sup-

lemented with 10% fetal bovine serum (FBS, Gibco). Full lengthDNAs of the HLA-DO� or HLA-DO� chains were obtained by PCRsing mRNA isolated from Raji cells. Initial cloning of the cDNAas performed using the pBlueScriptII KS vector and thereafter

logy 218 (2013) 1019– 1025

sub-cloned using the pEGFP-C2 or pDsRed-monomer-C2 vectors.Gene products were verified by sequencing. The produced HLA-DO� was fused at the amino-terminus with the fluorescent proteinEGFP, whereas HLA-DO� was fused at the amino-terminus with thefluorescent protein dsRed, facilitating thus its identification withinthe cellular compartments. The constructs shown in supplementFig. 1 were used to produce single or double transfectants of K562cells by electroporation. Stable transfectants were selected for aone-month period using the antibiotic G418 (1181131, Gibco). Alltypes of cells were submitted to immunofluorescence experimentsfollowed by flow cytometry analysis, whereas culture supernatantswere tested for the presence of free or exosome-engaged DR activ-ity by ELISA, transmission electron microscopy (TEM) or Westernblot analysis.

Supplementary material related to this article found, in theonline version, at http://dx.doi.org/10.1016/j.imbio.2012.12.003.

Exosome isolation and analysis

Exosomes were collected from the media of 4 ml K562 or HL-60 cells (2 × 106 cells/ml) cultured for 24 h in serum -free medium.Culture media was placed on ice and centrifuged at 800 × g for 6 minto sediment the cells and subsequently at 12,000 × g for 20 minat 4 ◦C to remove the cellular debris and concentrated to a vol-ume of 1 ml. The resulting supernatants were further centrifugedat 100,000 × g for 2 h at 4 ◦C. The exosome pellet was resuspendedeither in 20 �l paraformaldehyde 1% in phosphate buffer 0.2 M, pH7.4 for TEM analysis, or in loading buffer for SDS-PAGE analysisor in 200 �l of PBS for ELISA experiments. For the ELISA exper-iments, exosome membranes were osmotically disrupted. ELISA,SDS-PAGE and western blot analysis were performed as previ-ously described (Ranella et al. 2005; Athanassakis and Iconomidou,1996; Papadimitriou et al. 2008). In these experiments DR wasdetected using an uncoupled anti-HLA-DP-DQ-DR antibody. ForTEM analysis, exosomes were placed on Formvar carbon-coatedelectron microscopy grids, left to dry for 2 min, washed twice withPBS-BSA 0.1% and incubated with PBS-BSA 5% for 30 min at roomtemperature. After washing twice with PBS-BSA 0.1% grids wereincubated overnight with uncoupled anti-HLA-DP-DQ-DR at 4 ◦C,washed twice and developed using an anti-mouse IgG coupled to1.2 nm nanogold particles (4 h at room temperature). Nanogoldsignal was enhanced using the HQ SILVERTM enhancement kit(Nanoprobes NY, USA). After extensive washing, samples wereobserved using a JEOL-100C electron microscope equipped with aGATAN camera.

Sub-cellular fractionation

Sub-cellular fractionation was performed following the tech-nique described by Qiu et al. (1994). Briefly, 70 × 106 cells treated asdescribed above were collected, re-suspended in 2 ml of homoge-nization buffer (HB; 10 mM Tris, 1 mM EDTA, 0.25 M sucrose, pH7.4) and dounced gently in a Dounce Tissue Grinder (Wheaton357542; Wheaton Industries, Millville, NJ). The homogenate wascentrifuged at 900 × g and after collecting the supernatant in a cleantube, the pellet was resuspended in 1 ml HB and the 900 × g spinwas repeated. The resulting supernatant was combined with thefirst one and all were centrifuged at 10,000 × g to remove mitochon-dria. Two milliliters of the resulting supernatant were loaded ona 9 ml Percoll gradient (Pharmacia LKB Biotechnology Inc., Piscat-away, NJ; 1.05 g/ml) at 35,000 × g using a SW41 rotor (Beckman).

After collecting 0.5 ml samples from bottom to top, the resulting18–20 fractions were tested by ELISA for detection of DO activityand rab7. The results are expressed as percent of optical densityincrease over background values.

unobio

S

iwwe

R

L

6((hceda

ommCselac

mwhlEc(

Dc

Lc

mcerwsKKmmte

fc4sf

L. Papadimitriou et al. / Imm

tatistical analysis

Student “t” test was employed in order to compare the signif-cance levels (p) between control and test values. The KS analysis

as used for the flow cytometry obtained data. Correlation analysisas performed using the trendline options offered by the Microsoft

xcel application.

esults

ack of DO ̨ does not prohibit DO ̌ exit from ER

Previous studies had shown that the leukemic K562 and HL-0 cells lack DO� mature transcripts while expressing DO�Papadimitriou et al. 2008). In H2-Oa-deficient mice, Liljedahl et al.1998) postulated that in the absence of H2-Oa, the H2-O� chainad a decreased half-life and did not associate with H2-M, indi-ating that it was not transported out of the ER. However, Samaant al. (1999), using various HeLa DO� transfectants were able toetect DO� in lysosomal compartments suggesting a stable DO–DMssociation.

In order to evaluate whether DO� exits ER in the naturallyccurring DO�−�+ K562 and HL-60 cells, confocal and electronicroscopy analysis as well as sub-cellular fractionation experi-ents were performed using an anti-DO� (clone DOB.L1) antibody.

onfocal microscopy analysis co-localized DO� with the rab7 endo-omal marker in both K562 and HL-60 cells (Fig. 1A). Transmissionlectron microscopy experiments which can clearly distinguish cel-ular compartmentalization showed in HL-60 cells that DO� waslways detected outside the ER membranes (Fig. 1B; four differentellular areas are shown).

Following sub-cellular fractionation protocols, the endoso-al/lysosomal vesicles were separated through a Percoll gradient,here endosomes to lysosomes could be isolated from light toeavy fractions, respectively. Upon membrane disruption the iso-

ated fractions were tested for the presence of DO� and rab7 byLISA (Fig. 1C). The results showed that in both K4562 and HL-60ells, DO� co-localized with rab7 in the light endosomal fractionsFig. 1C).

These findings demonstrated that in the naturally occurringO�−�+ K562 and HL-60 cells, DO� accumulates in the endosomalompartments.

ack of DO ̨ is partially responsible for the absence of DR from theell membrane.

As previously shown K562 and HL-60 cells do not expressembrane DR (mbDR) but retain these molecules intra-cellularily,

omplexed to DO� in the endosomal compartments (Papadimitriout al. 2008). In order to test whether the absence of DO� wasesponsible for arresting DR into endosomes, DO� (DO�-dsRed)ith or without DO� (DO�-EGFP) were introduced in K562 and

table transfectants were obtained namely, K-DO�, K-DO� and-DO�-DO�. Flow cytometry analysis showed that K-DO� and-DO�-DO� but not K-DO� displayed an approximately 20% ofbDR expression as evaluated using an anti-HLA-DR-FITC-labeledouse anti-human antibody (Fig. 2A). These results indicated that

he absence of DO� was partially responsible for DR arrest in thendosomal compartments.

Introduction of DO� could rescue some mbDR in K-DO� trans-ectants (5-fold increase of mbDR expression as compared to

ontrol K562 cells; 13–17% of expression), while increasing almost-fold DM expression (Fig. 2B). Thus the low levels of DM expres-ion in K562 cells (2–5% of expression) could be a limiting factoror further mobilization of DR to the cell membrane. To test this

logy 218 (2013) 1019– 1025 1021

hypothesis, K-DO� transfectants were treated with IFN-� for 24 hand tested for mbDR expression. Flow cytometry analysis showedthat this treatment increased mbDR expression to 50% (2-fold ascompared to untreated cells), while decreasing intracellular DRexpression (Fig. 2B). The results indicated that both DO� and DMshould be responsible for the intracellular arrest of DR.

DO ̌ drives DR to exosomes

As mentioned above, K562 and HL-60 cells express only lowlevels of DM (expression varying from 2–5 and 2–8% in K562 andHL-60 cells, respectively; data not shown).

Using multiple transfectants of HeLa cells, Xiu et al. (2011) havepostulated that DO limits DM availability in exosomes and that DO�could bind to DR and facilitate transport to exosomes. In accordanceto these observations, it could be hypothesized that in K562 and HL-60 cells, DO� in association with the low levels of DM could driveat least a portion of intracellular DR to exosomes.

To test the above hypothesis, co-expression of intracellular DRand the exosome marker CD9 was evaluated in K562 and controlRaji cells. Following flow cytometry analysis it was shown that theDR:CD9 ratio was 0.45 and 0.79 in K562 and Raji cells respectively(Fig. 3A), which is in agreement with the constitutive expression ofmbDR in Raji cells.

To further evaluate exosome secretion from K562 and HL-60 cells, 24-hour supernatants from serum-free cultures in thepresence or not of IFN-� were submitted to exosome isolationprocedures. Exosomes and the exosome-free fraction were testedfor DR activity by ELISA (Fig. 3B), Western blot (Fig. 3C) andtransmission electron microscopy analysis (Fig. 3D). Followingthese experimental protocols DR activity could be detected inboth exosome and exosome-free fractions, indicating that multiplemechanisms of excretion can drive DR to the extracellular matrix.The presence of IFN-� significantly increased exosome-engagedand exosome-free DR protein only in K562 cells (p < 0.005; data notshown). Western blot analysis of exosomes, detected a ∼60 kD�band in all cases tested (Fig. 3C), which is in agreement with theprotein size of the soluble MHC II isolated from serum (Sardis et al.2010). The presence of DR in exosomes could also be detected byimmunostaining followed by silver enhancement and TEM analysis(Fig. 3D).

In order to test whether DO� could alter DR excretion, 24 hsupernatants from K562 transfectants were tested for the pres-ence of DR protein. Indeed, K-DO� and K-DO�� but not K-DO�transfectants displayed reduced DR secretion by 2.4-fold as com-pared to control K562 cells (p < 0.001; Fig. 3E1). Interestingly, DRsecretion was shown to negatively correlate to mbDR expression(y = −0.0329x + 23.385; R2 = 0.9873; Fig. 3E2) as well as DM expres-sion (y = −0.3772x + 62.287; R2 = 0.8982; data not shown). Theseresults indicated that in the absence of DO�, DO� could drive DRto excretion through exosomes. Whether DO� is also responsiblefor DR exosome-free secretion remains to be explored.

Discussion

Maturation of class II MHC antigens, antigenic epitope loading,transport to the cell membrane and recognition by CD4-positivecells constitutes one of the major arms of the immune response,leading to antigen recognition and attack. Most leukemic cells failto express classical DR molecules and are therefore prone to escapeimmune surveillance. Studying the leukemic cell lines K562 and HL-

60, one can extract information not only at the level of mechanismsgoverning immune escape of malignant cells, but also the regula-tory pathways of DR maturation procedures in antigen presentingcells. The work presented here showed the lack of DO� and the

1022 L. Papadimitriou et al. / Immunobiology 218 (2013) 1019– 1025

Fig. 1. Expression of HLA-DO in K562 and HL-60 cells. Immunofluorescence staining followed by confocal microscopy analysis (A) reveals HLA-DO, rab7 as well as the co-localization of HLA-DO/rab7. Only a Z-slice of the cells is shown here. Electron microscopy analysis of HL-60 cells (B) shows the expression of HLA-DO at 4 different cellularareas. Magnification of 50,000× (bars 0.4 �m). (M = mitochondria, ER = endoplasmic reticulum). Localization of DO and rab7 in the endosomal/lysosomal compartments ofK562 and HL-60 cells was evaluated by ELISA of subcellular fractions (C) as described in the Methods section. The fractions have been collected from bottom-to-top of thedensity gradient and represent the lysosomal-to-endosomal structures. The results are expressed as the percent of optical density readings over background levels andrepresent the mean of triplicate samples of one out of 3 experiments performed in each case, where similar results were obtained. SEM varied in all cases from 3 to 6%.

L. Papadimitriou et al. / Immunobiology 218 (2013) 1019– 1025 1023

Fig. 2. Membrane DR expression in control and K562 transfectants. Various K562 transfectants (Supplement Fig. 1) were tested for surface DR expression by immunofluo-rescence, followed by flow cytometry analysis (A). Modulation of DR expression in K-DO� by IFN-� (B). K-DO� transfectants were treated with IFN-� for 24 h and examinedf escenfl f 3 ex

lftdr

nkbaees

or the expression of membrane or intracellular/total DR and DM by immunofluorow cytometry analysis are shown (C). In all cases the results represent the mean o

ow expression of DM could be responsible for the absence of DRrom the cell membrane and that DO� could drive DR moleculeso exosomes for excretion. The use of various transfectants allowsefining novel regulatory pathways, providing new insights in theegulation of class II antigen maturation.

The results presented here showed that the lack of DO� didot prohibit DO� exit from ER. Since DO� does not contain anynown ER-retention motifs in the short cytoplasmic tail, it haseen postulated that, at least in B cells, DM acts as a chaperone,

ssisting � � pairing to prevent degradation of HLA-DO (Brunett al. 2000). However, in the absence of DO� and the low levelxpression of DM, it was shown that DO� could enter the endo-omal compartments. Confocal microscopy and transmission

ce followed by flow cytometry analysis (B), whereas representative data from theperiments and are expressed as percent of positive cells (±SEM).

electron microscopy analysis, as well as sub-cellular fractionationexperiments localized DO� outside ER, co-localizing with rab7in endosomes. These results are in accordance with previousobservations, where DO� was shown to form stable complexeswith DR in endosomes (Papadimitriou et al. 2008). Indeed, thedi-leucine sorting signal in the cytoplasmic tail of DO� allowsthese molecules to freely traffic into endosomes (Xiu et al. 2011).

In order to evaluate whether the lack of DO� was responsi-ble for prohibiting DR migration to the cell membrane, DO� with

or without DO� was introduced to K562 cells and examined formbDR expression. The results showed that stable K-DO� and K-DO�DO� transfectants expressed mbDR but at levels not exceeding20% of expression. In order to test whether DM could also be a

1024 L. Papadimitriou et al. / Immunobiology 218 (2013) 1019– 1025

Fig. 3. Excretion of DR proteins in a free and exosome-engaged form. Co-expression of intracellular DR and CD9 in K562 and Raji cells was evaluated by flow cytometry analysis(A). The secretion of free or exosome-engaged DR proteins in K562 and HL-60 cells, isolated as described in methods, was evaluated by ELISA (B). The results are expressed aspercent of optical density increase over background (±SEM) and represent the mean of 5 experiments. The DR content in exosomes isolated from 24 h supernatants of control(−) or IFN-�-treated (+) K562 or HL-60 cells was also evaluated by PAGE gel (left panel; M: molecular weight marker) followed by western blot analysis (right panel) (C). Thelanes were loaded with 40 �l sample (32 �l test sample mixed with 8 �l of loading buffer −0.025 M Tris, 5% �-mercaptoethanol, 2% SDS, 20% glycerol, 0,1% bromoethanolblue) and represent (1): broken exosomes (ultra-centrifugation-derived pellet-exosomes-resuspended in 40 �l PBS), (2): exosome-free culture supernatant (supernatantsrecovered from the ultra-centrifuged samples), and (3): total culture supernatant (concentrated supernatants before ultra-centrifugation). Exosomes isolated from K562or HL-60 cells were also submitted to immunostaining and TEM analysis (D). Exosomes were stained with an anti-HLA-DR antibody, visualized using an anti-IgG coupledto 1.2 nm nanogold particles, followed by signal enhancement using the HQ SILVERTM enhancement kit. Total DR protein detection in 24 h supernatants of K-DO� and/orK-DO� transfectants, was evaluated by ELISA (E1). In this case, 4x-concentrated supernatants were submitted to freezing/thawing for exosome membrane disruption. Ther ) and

i

l�wDtmd

osswasE

esults are expressed as percent of optical density increase over background (±SEMn control K562 and K-DO� and/or K-DO� transfectants (E2).

imiting parameter, K-DO� transfectants were treated with IFN-. Indeed, such treatment increased mbDR expression by 2,5-fold,hile also increasing the levels of DM. These results indicated thatO� and DM were the limiting factors, disallowing DR to reach

he cell membrane. Indeed a positive correlation between DM andbDR (y = 1.2643x − 1.7582; R2 = 0.9631; data not shown) could be

etected.Although lacking DO�, K562 and HL-60 cells express high levels

f DO�, which could be responsible for driving DR intro exo-omes (Xiu et al. 2011). Indeed, double fluorescence experimentshowed that DR was co-expressed with the exosome marker CD9,

hile ELISA experiments detected free and exosome-engaged DR

ctivity in the culture supernatants, with the precaution that theo-called free DR protein could derive from broken exosomes.xosome-engaged DR excretion was also detected by EM imaging

represent the mean of 3 experiments. Secreted DR was correlated to mbDR protein

and western blot analysis. Soluble MHCII antigens loaded with selfpeptides have been considered to maintain tolerance (Fischer et al.2007). Such mechanism could function in favor of malignant cells,where soluble class II molecules could negatively regulate T cellactivation to avoid immune attack (Sardis et al. 2010; Vevis et al.2012). Furthermore, the presence of DO� in K-DO� and K-DO��transfectants significantly reduced DR excretion, which was shownto inversely correlate to mbDR expression in these cells.

In conclusion, it seems that the lack of DO�, the high DO� andlow DM expression levels in K562 and HL-60 cells are responsiblefor prohibiting DR migration to the cell membrane, while driving

these molecules to exosome-engaged or exosome-free excretion.Such mechanism could act in favor of leukemic cell survival andescape from immune attack, where soluble DR molecules couldfurther suppress T cell activation and allow leukemic cell survival.

unobio

Acptla

A

aatCcc

R

A

B

C

C

D

D

F

Wake, C.T., Flavell, R.A., 1985. Multiple mechanisms regulate the expression ofmurine immune response genes. Cell 42, 623–628.

Xiu, F., Côté, M.H., Bourgeois-Daigneault, M.C., Brunet, A., Gauvreau, M.É., Shaw, A.,

L. Papadimitriou et al. / Imm

lthough the presented results provide novel insight in intra-ellular class II MHC trafficking pathways, adding a few moreieces to the antigen presentation puzzle, it would be very impor-ant to examine whether such mechanisms also apply in primaryeukemias, where novel therapeutic approaches could be envis-ged.

cknowledgements

This work was supported by the General Secretariat for Researchnd Technology (75%) and the European Community State (25%),s well as the national funding of the Ministry of Education andhe Special Account for Research Resources of the University ofrete through services to the private sector. We thank the Spe-ial Account for Research Resources of the University of Crete forovering publication costs.

eferences

thanassakis, I., Iconomidou, B., 1996. Cytokine production in the serum and spleenof mice from day 6 to 14 of gestation. Dev. Immunol. 4, 247–255.

runet, A., Samaan, A., Deshaies, F., Thomas, J., Kindt, T.J., Thibodeau, J., 2000. Func-tional characterization of a lysosomal sorting motif in the cytoplasmic tail ofHLA-Dob. J. Biol. Chem. 275, 37062–37071.

hamuleau, E.D., Souwer, T., van Ham, S.M., Zevenbergen, A., Westers, T.M., Berkhof,J., Meijer, C.I., van de Loosdrecht, A.A., Ossenkoppele, G.J., 2004. Class II-associated invariant chain peptide expression on myeloid leukemic blastspredicts poor clinical outcome. Cancer Res. 64, 5546–5550.

hen, X., Laur, O., Kambayashi, T., Li, S., Bray, R.A., Weber, D.A., Karlsson, L., Jensen,P.E., 2002. Regulated expression of human histocompatibility leukocyte antigen(HLA)-DO during antigen-dependent and antigen-independent phases of B celldevelopment. J. Exp. Med. 195, 1053–1062.

enzin, L.K., Sant’Angelo, D.B., Hammond, C., Surman, M.J., Cresswell, P., 1997. Nega-tive regulation by HLA-DO of MHC class II-restricted antigen processing. Science278, 106–109.

enzin, L.K., Fallas, J.L., Prendes, M., Yi, W., 2005. Right place, right time, right pep-

tide: do keeps DM focused. Immunol. Rev. 207, 279–292.

ischer, U.B., Jacovetty, E.L., Medeiros, R.B., Goudy, B.D., Zell, T., Swanson, J.B., Lorenz,E., Shimizu, Y., Miller, M.J., Khoruts, A., Ingulli, E., 2007. MHC class II deprivationimpairs CD4 T cell motility and responsiveness to antigen-bearing dendritic cellsin vivo. Proc. Natl. Acad. Sci. USA 104, 7181–7186.

logy 218 (2013) 1019– 1025 1025

Kropshofer, H., Vogt, A.B., Moldenhauer, J., Hammer, J., Blum, J.S., Hammerling,G.J., 1996. Editing of the HLA-DR-peptide repertoire by HLA-DM. EMBO J. 15,6144–6154.

Liljedahl, M., Winqvist, O., Surh, C.D., Wong, P., Ngo, K., Teyton, L., Peterson, P.A.,Brunmark, A., Rudensky, A.Y., Fung-Leung, W.P., Karlsson, L., 1998. Altered anti-gen presentation in mice lacking H2-O. Immunity 8, 233–243.

Morris, P., Shaman, J., Attaya, M., Amaya, S., Goodman, C., Bergman, J.J., Monaco, J.J.,Mellins, E., 1994. An essential role for HLA-DM in antigen presentation by classII major histocompatibility molecules. Nature 368, 551–554.

Papadimitriou, L., Morianos, I., Michailidou, V., Dionyssopoulou, E., Vassiliadis,S., Athanassakis, I., 2008. Characterization of intracellular HLA-DR, DMand DO profile in K562 and HL-60 leukemic cells. Mol. Immunol. 45,3965–3973.

Qiu, Y., Xu, X., Wandinger-Ness, A., Dalke, D.P., Pierce, S.K., 1994. Separation of sub-cellular compartments containing distinct functional forms of MHC class II. J.Cell Biol. 125, 595–605.

Ranella, A., Vassiliadis, S., Mastora, C., Michailidou, V., Dionyssopoulou, E., Athanas-sakis, I., 2005. Constitutive intracellular expression of HLA-DO and HLA-DR butnot HLA-DM in trophoblast cells. Hum. Immunol. 66, 43–55.

Samaan, A., Thibodeau, J., Mahana, W., Castellino, F., Cazenave, P.A., Kindt, T.J., 1999.Cellular distribution of a mixed MHC class II heterodimer between DRa and achimeric DOb chain. Int. Immunol. 11, 99–111.

Sardis, M.F., Miltiadou, P/., Bakela, K., Athanassakis, I., 2010. Serum-derived MHCclass II molecules: potent regulators of the cellular and humoral immuneresponse. Immunobiology 215, 194–205.

Tonnelle, C., DeMars, R., Long, E.O., 1985. DO�: a new � chain gene in HLA-D withdistinct regulation of expression. EMBO J. 4, 2839–2847.

vam Ham, S.M., Tjin, E.P.M., Lillemeir, B.F., Gruneberg, U., van Maijgaarden, K.E.,Pastoors, L., Verwoerd, D., Tulp, A., Canas, B., Rahman, D., Ottenhoff, T.H., Pappin,D.J., Trowsdale, J., Neefjes, J., 1997. HLA-DO is a negative regulator of HLA-DM-mediated MHC class II peptide loading. Current Biol. 7, 950–957.

vam Ham, S.M., van Lith, M., Lillemeir, B.F., Tjin, E., Gruneberg, U., Rahman, D., Pas-toors, L., van Meijgaarden, C., Roucard, J., Trowsdale, J., Ottenhoff, T., Pappin,D., Neefjes, J., 2000. Modulation of the major histocompatibility complex class IIassociated repertoire by human histocompatibility leukocyte antigen (HLA)-DO.J. Exp. Med. 191, 1127–1136.

Vevis, C., Matheakakis, A., Kyvelidou, C., Bakela, K., Athanassakis, I., 2012. Charac-terization of Antigen-binding and MHC class II-bearing T Cells with SuppressiveActivity in Response to Tolerogenic Stimulus. Immunobiology 217, 100–110.

Thibodeau, J., 2011. Cutting edge: HLA-DO impairs the incorporation of HLA-DMinto exosomes. J. Immunol. 187, 1547–1551.